Circuits with non-linear characteristics are present in many electronic systems. The transistor, the ubiquitous fundamental component of modern electronic systems, exhibits non-linear characteristics. The linear amplifier is referred to as “linear” merely to refer to operating conditions and/or the circuit design aspects that minimize its non-linear effects. For example, the linear amplifier as implemented in high-speed communication (e.g., fiber optic) networks and systems is expected to exhibit accurate (e.g., high linearity, low distortion) amplification of the input data signals over a maximum output signal range. Distortion and/or other non-linear characteristics introduced by the linear amplifier can result in an increase in bit-error-rate (BER) and/or other system performance issues. In some cases, a limited output signal range can limit the ability of the fiber optic system to implement higher orders of multi-level signaling (e.g., PAM-16 having 16-level signaling) to accommodate the continually increasing demand for more data bandwidth over a fixed transmission medium. Designers of linear amplifiers and other circuits with non-linear characteristics use circuit simulators to evaluate the behavior of their circuits under various conditions (e.g., simulation constraints) so as to predict the performance of the circuit in its target environment (e.g., a high speed communication system).
Unfortunately, legacy approaches to simulating non-linear behavior in circuits can present limitations at least as pertaining to efficiently creating simulation scenarios that facilitate accurate predictions of non-linear circuit behavior. Specifically, certain legacy approaches might select a simulation input signal characterized by a single tone (e.g., one frequency) or a dual tone (e.g., two frequencies). In such cases, the simulation time can be short, but the simulation results can be limited. Specifically, a single tone simulation can produce a prediction of total harmonic distortion (THD). However, the THD metric might not be correlated to system performance. Further, a dual tone simulation can produce a prediction of intermodulation distortion (IMD), but the IMD metric might also be limited as to its correlation to system performance. Some legacy approaches use modulated simulation signals to predict an error vector magnitude (EVM) of the circuit. For example, a quadrature amplitude modulation (e.g., 16 QAM) signal corresponding to the end application of the circuit can be used for simulation. However, the simulated EVM of the circuit might be limited in its correlation to the EVM of the system due to such factors as equalization (e.g., amplitude and phase compensation) included in the system. Further, the duration of EVM simulations can be long. Some legacy approaches might use multitone simulation signals. However, selecting the set of tones and other signal attributes (e.g., amplitude, phase, etc.) to accurately predict the performance of the circuit in the target system can be limited.
Techniques are therefore needed to address the problem of efficiently selecting a simulation input signal to accurately predict non-linear characteristics of a circuit as implemented in an electronic system. None of the aforementioned legacy approaches achieve the capabilities of the herein-disclosed techniques for dynamic multitone simulation scenario planning. Therefore, there is a need for improvements.